1. Field of the Invention
The invention relates to a coding method for high-efficiently coding an image signal and an image signal coding apparatus for realizing the method.
2. Description of the Related Art
The industries over the world have conventionally been discussing how to high-efficiently code an image signal. The high-efficiently coding methods have been standardized by the Moving Picture Experts Group (MPEG), which is one section of the joint technology committee of the ISO (International Organization for Standardization) and the IEC (International Electro-technical Commission). This standardization brought about such known results as ISO/IEC 13818-2 “Coding of Audio, Picture, Multimedia, and Hypermedia Information” (MPEG-2 Standard) and ISO/IEC DIS 11172 “Coding of Moving Pictures and Associated Audio ISO/IEC JTC1/SC29 WG11” (MPEG-1 Standard).
To improve the picture quality of an image signal which is high-efficiently coded on the basis of these standards, there is available a method for once coding an image signal under predetermined coding conditions and then re-coding it corresponding to its characteristic. A technology related to this method is disclosed in Japanese Unexamined Patent Publication No. Hei 9-214962 (1997) gazette.
The following will describe a conventional image signal coding method with reference to FIG. 5. A coding means 501 codes an image signal. A parameter setting means 502 sets coding conditions. An encode information creating means 503 extracts a characteristic from an input image signal based on a coded result. An allocation means 504 calculates optimal coding conditions based on the information of a characteristic extracted from the image signal.
The operations are as follows. An input image signal is coded by the coding means 501 according to a predetermined initial parameter set by the parameter setting means 502. The result obtained by coding and the conditions on coding are sent to the encode information creating means 503 to be analyzed there. The encode information creating means 503 determines coding difficulty of each picture constituting the input image signal by referencing such information as code amount generated on coding, quantization steps used for coding, magnitude of a motion vector, an activity, execution or non-execution of scene change and the like.
The analysis result obtained by the encode information creating means 503 is sent to the allocation means 504. The allocation means 504 calculates data distribution to each of the pictures based on respective coding difficulty of each of the pictures created by the encode information creating means 503. According to the data distribution calculated by the allocation means 504, the parameter setting means 502 changes the coding conditions, so that the coding means 501 re-codes the input image signal based on thus changed new coding conditions.
By the operation above, the input image signal can be coded corresponding to its characteristic. The coded picture quality can be improved as well. In case that a satisfactory coded picture quality cannot be obtained by executing coding only once, the processing steps mentioned above are repeated any number of times in a loop. Each time the processing is thus repeated, the picture quality is improved further. In this case, rather than repeating all the steps of the processing, only the calculation part by the allocation means 504 may be repeated any number of times in the loop, thus optimizing data distribution.
The method for image signal coding mentioned above takes a long time to carry out because the allocation means 504 must calculate data distribution for the whole of an input image signal. The amount of parameters is also huge which is set on coding required to carry out this data distribution. To realize this method, an associated apparatus becomes large in scale and hence in cost. To improve the picture quality, the processing mentioned above must be repeated a plurality of times in a loop, which in turn increases the processing time.